IRFinder: assessing the impact of intron retention on mammalian gene expression

Abstract

Intron retention (IR) occurs when an intron is transcribed into pre-mRNA and remains in the final mRNA. We have developed a program and database called IRFinder to accurately detect IR from mRNA sequencing data. Analysis of 2573 samples showed that IR occurs in all tissues analyzed, affects over 80% of all coding genes and is associated with cell differentiation and the cell cycle. Frequently retained introns are enriched for specific RNA binding protein sites and are often retained in clusters in the same gene. IR is associated with lower protein levels and intron-retaining transcripts that escape nonsense-mediated decay are not actively translated.

Keywords: Gene regulation; Intron retention; mRNA splicing.

Fig. 1

Intron retention is widespread and its measurement is consistent within the same tissue type. a Number of retained introns in each of the 2573 downloaded samples that were retained in more than 10% (red), 20% (green), 30% (blue) of transcripts (IR ratio >0.1, 0.2, 0.3). b Average IR ratio in eight different tissue types. hESC human embryonic stem cell. c Genes with IR events occur closer than expected. Genomic distance between genes with IR events (blue) compared to a control of the same number of randomly selected expressed genes. d Retained introns within the same gene are often adjacent to each other. Number of introns that are separated by only one exon (x-axis) in the same gene (blue) compared to a control where the number of IR events within each gene are randomly distributed amongst introns

Fig. 2

IR events alter protein output. a Genes with IR have significantly lower protein output (p values are from multivariate anova comparing mRNA and protein levels between IR and non-IR). Protein and mRNA levels normalized using a standard transformation for nine tissue types. A lowess regression curve shows the correlation between protein and mRNA levels. Genes with IR events are colored according to the maximum IR ratio of retained genes. This ratio reflects the percentage of transcripts that retain introns for a specific gene. b IGV screenshots of CHX Ribo-seq, QTI-seq, and mRNA-seq. Gene structure is displayed below each graph with horizontal lines indicating introns and blocks indicating exons. The sequencing coverage (number of reads) is displayed to the left. IR regions are highlighted in green

Fig. 3

IR marks stages of cell differentiation and the cell cycle. a IR measured during 4 days of overexpression of Neurogenin-1 and -2 to differentiate human embryonic stem cells (hESCs) into bipolar neurons [16]. b IR and mRNA expression of two crucial genes, SOX15 and POU4F1. c Heatmap of IR measured during differentiation of fibroblasts from genetically matched hESC and human induced pluripotent stem cells (hiPSC) lines. Left: hESCs were grown in extended culture then differentiated into fibroblasts. Right: hESCs are differentiated to fibroblasts, then dedifferentiated to hiPSCs and then differentiated again into fibroblasts. Curly brackets indicate replicates. d Gene Ontology (GO) functional category enrichment after Benjamini–Hochberg (BH) correction. EGF epidermal growth factor. e IR measured for TH2 cells sorted at different stages of the cell cycle. f GO functional category enrichment after BH correction

Fig. 4

RNA binding proteins (RNAbps) modulate IR. a RNAbp motifs enriched (Z score >1.96) in frequently retained introns (5′ and 3′ intron boundaries) and flanking exons. No depleted motifs were found. b Number of increased and decreased IR events at different p values of significance following knockdown (KD). Total numbers of increased and decreased IR events are indicated on the plot. Motif enrichment p values in increased IR events for each RNAbp are displayed below their protein names. c The number of increased IR events compared to the number of retained introns following knockdown for which there was no evidence of retention before knockdown

Fig. 5

Overview of the IRFinder algorithm. QC quality control. ERCC Sequences from External RNA Controls Consortium (ERCC, http://jimb.stanford.edu/ercc/)